Evodevo and the Promise of Understanding Morphological
Transitions in Evolution
Author(s): Lisa M. Nagy and Terri A. Williams
Source: Annals of the Missouri Botanical Garden, 99(3):289-300. 2014.
Published By: Missouri Botanical Garden
DOI: http://dx.doi.org/10.3417/2011036
URL: http://www.bioone.org/doi/full/10.3417/2011036
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the
biological, ecological, and environmental sciences. BioOne provides a sustainable
online platform for over 170 journals and books published by nonprofit societies,
associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content
indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/
page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and noncommercial use. Commercial inquiries or rights and permissions requests should be
directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit
publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to
critical research.
Volume 99
Number 3
2014
Annals
of the
Missouri
Botanical
Garden
EVODEVO AND THE PROMISE OF
UNDERSTANDING
MORPHOLOGICAL TRANSITIONS
IN EVOLUTION1
Lisa M. Nagy2 and Terri A. Williams3
ABSTRACT
The field of Evolutionary Developmental biology arose with the promise of new approaches to answering longstanding
questions of comparative biology. Here we review the fruits of that promise some decades later. We chose three areas of
arthropod EvoDevo—evolution of body plans, segment number, and appendage morphology—to provide an overview for the
nonspecialist of how these issues have been clarified by the comparative analysis of regulatory gene networks. In all cases, we
identify substantial progress and novel insights provided by the tools and perspective of EvoDevo. We also recognize that some
core questions remain unanswered, and we reflect on how discoveries in EvoDevo fit in the landscape of other progress in
phylogenetics, population biology, and genomics, facilitated by a new and ever-expanding set of molecular tools for comparative
studies in evolution.
Key words: Appendage development, arthropods, development, evolution, segmentation.
What meets the eye when we cursorily inspect
nature is an overwhelming variety of morphological
forms. One current strategy in biology to explain how
diverse forms might have evolved is to compare the
regulation of body patterning during development. If
we can grasp how form develops among a number of
related species, we can hypothesize how modifications in development create distinct morphological
forms over evolutionary time. The contemporary study
of how developmental patterning evolves—EvoDevo
—relies primarily on understanding the gene regulatory pathways that modulate development. At the
same time, EvoDevo draws on longstanding intellec-
tual enterprises in science. Here, we briefly describe
the development of EvoDevo as a modern field. Then,
using three specific examples related to our research
in arthropods, we evaluate the success of this
approach.
One of the oldest insights into animal diversity is
that variety can be partitioned and comprehended by
grouping similar animals together. Discriminating
similarities and differences among animals and using
those to erect categories of distinct types of animals
goes back at least to Aristotle and was a continuing
thread in the natural sciences as they developed over
the next two millennia. By the late 18th to early 19th
1 This and the following three articles are the proceedings of the 57th Annual Systematics Symposium of the Missouri
Botanical Garden, ‘‘EVO-DEVO: Where Are We Now, Where Are We Going?’’ The symposium was held 15–17 October
2010, at the Missouri Botanical Garden in St. Louis, Missouri, U.S.A.
2 Molecular and Cellular Biology, University of Arizona, Tucson, Arizona 85721 U.S.A. [email protected].
3 Department of Biology, Trinity College, Hartford, Connecticut 06106 U.S.A. [email protected].
doi: 10.3417/2011036
ANN. MISSOURI BOT. GARD. 99: 289–300. PUBLISHED
ON
15 MAY 2014.
290
Annals of the
Missouri Botanical Garden
Figure 1. Examples of relationships of body plans discerned from morphological and molecular perspectives. —A. Geoffroy
Saint-Hilaire’s famous drawing of a lobster dissection ‘‘une coupe longitudinale du homard’’ from plate 7 (p. 119) of the 1822
article ‘‘Considérations générales sur la vertèbre.’’ In our Figure 1, the lobster is shown lying on its back, with its ventral nerve
cord above the internal organs. In the inverted orientation herein, the body plan of the arthropod resembles that of the vertebrate.
—B. Diagrammatic views of the protostome and chordate body plans. The dorsoventral structures occupy opposite sides of the
body and are patterned by inverted domains of the diffusible growth factors (green gradient: dpp/BMPs) and their inhibitors (red
gradient: sog/Chordin; from DV-axis-inversion, L’ontogenese, Wikipedia). —C. Generalized pattern of arthropod tagmatization.
The diagram shows a simplified representation of the patterns of segmental diversification within arthropods: changes in segment
number, presence or absence of segments on a particular segment, and specialization of appendages within one body region.
century, the conceptual framework that developed for
this enterprise used the idea of homology (a structure
similar under any transformation) versus analogy (any
structure of similar function) to interpret parts of
animals, and the similarities among animals were
generalized using the conceptual model of an
archetype (see Russell, 1917; Hall, 1994, 1999).
The use of archetypes was a powerful tool for making
comparisons highly explicit between taxa since the
archetype was essentially a series of hypotheses
about the morphology of a particular taxon. This
theoretical framework, informed by a sophisticated
grasp of the body plans of distinct taxa, led to specific
discoveries (e.g., Goethe’s discovery of the human
intermaxillary bone) as well as broad sweeping
theories (e.g., Geoffroy Saint-Hilaire’s theory that
vertebrates are essentially arthropods flipped onto
their backs, or Richard Owen’s demonstration via his
vertebrate archetype that vertebrates are built from
repeated parts (Fig. 1A; Russell, 1917; Appel, 1987;
Hall, 1999).
The theoretical framework of comparative morphology underwent a radical transformation at the
end of the 19th and early 20th century due to two
developments: Darwin’s theory of evolution by natural
selection and the modern synthesis of Mendelian
genetics and population genetics (see Mayr, 1993;
Gilbert et al., 1996; Bowler, 2003). The growth and
predominance of these ideas had the effect of
diverting the understanding of morphology from
comparisons of form to a search for genetic (or other
reductive) causes of form. In this new light, many of
the hypotheses of the previous century disappeared.
The conundrums of the past century were not
resolved; they simply became unimportant in the
new way of conceiving of natural phenomena.
However, while old theories and hypotheses based
on archetypes were not part of the modern research
Volume 99, Number 3
2014
Nagy & Williams
Morphological Transitions in Evolution
291
agenda, they were still embedded in most textbooks
dealing with animal diversity. For example, textbooks
still presented generalized schemata to illustrate
different phyla (see Brusca & Brusca, 2003) with
traits of those schemata clearly defined (so-called
true segmentation vs. pseudosegmentation vs. no
segmentation). The persistence of these schemata was
important to the emergence of EvoDevo as a field
because they provided the background for comprehension of the importance of different fundamental
body plans to metazoan diversity.
Developmental genetics made rapid progress in
understanding the genetic regulatory control of
external morphology in a few select model organisms.
EvoDevo as a field gained momentum with the
discovery that many of the genetic regulatory
mechanisms that drive patterning in model organisms
were broadly shared. A commonly cited example is
Pax6, a gene for a transcription factor used to
position the eyes in diverse organisms, even eyes that
are not homologous (reviewed in Gehring, 2002).
Carroll et al. (2001) postulated that a finite genetic
toolkit existed for patterning embryos, and following
on earlier ideas that changes in gene regulation were
critical for phenotypic diversity (Jacob, 1977),
popularized the idea that tinkering with this
conserved set of regulatory genes could produce
diverse body plans. This hypothesis reanimated some
of the old questions about how the basic body plans of
animals relate to one another. For example, a
regulatory loop that patterned the dorsal axis of
arthropods, but the ventral axis in vertebrates,
resurrected Geoffroy’s theory of vertebrates being
upside-down arthropods (Fig. 1B; De Robertis &
Sasai, 1996). From a gene’s eye view, the genetic
basis underlying morphological variation appeared to
be remarkably similar throughout metazoans (see
Carroll, 2005). This led to an initial enthusiasm that a
modern, gene-based EvoDevo would resolve old
issues of comparative morphology and ultimately be
able to explain the great morphological radiations.
Here, we provide a brief review of progress on
EvoDevo as viewed through the lens of examples
relevant to our research dealing with the diversification of arthropods. In arthropods, a segmented body
plan covered by a chitinous exoskeleton has
produced evolutionary radiations of highly diverse
and elaborated external morphologies. Many body
segments bear appendages, and these are often
specialized to perform distinct functions, both along
the body axis in any particular species as well as
between species. Much of arthropod diversity can be
ascribed to segments and their appendages. In the
vignettes that follow, we reflect on the power of the
EvoDevo approach to shed light on that diversity, and
we conclude with reflections on why some questions
have proved more tractable than others and what
future approaches might include. We examine three
features of variation of the arthropod body plan: (1)
tagmatization and limb character along the anteriorposterior (A-P) body axis, (2) segment number, and
(3) limb morphology. In each case, we outline the
variation to be explained, the hypotheses generated
from the genetic model system, the existing data, and
what they explain.
TAGMATIZATION
VARIATION TO BE EXPLAINED
Arthropod diversity can be grossly characterized
by variation in the numbers and specializations of
segments. Most taxa have a fixed total number of
segments, but some branchiopod crustaceans and
some centipedes have a varying total number of
segments. A feature common to all taxa is tagmatization, the regionalization of the body into distinct
blocks of segments, namely, the head, thorax, and
abdomen (Fig. 1C). Not only does total segment
number vary between species, but also the number of
segments in any particular body region can vary
among taxa, e.g., the insect thorax has three
segments, the decapod crustacean thorax has eight.
Furthermore, overlaid on this divergence in tagma
between distantly related taxa are modifications to
tagma among even closely related taxa. For example,
among decapods with eight thoracic segments (e.g.,
crabs, shrimp, lobsters) some might have all eight
specialized for locomotion while others divide the
eight into functional subspecialties. A common
example is the appendages on the anterior thoracic
segments that are modified to function with the head
segments in feeding.
In general, the variation in arthropod segmentation
can be grouped into three categories: (1) whether the
total number of body segments is variable or fixed; (2)
if it is fixed, how one taxon differs from another in
total segment number and tagmatization; and (3) how
body regions within a taxon are modified. These types
of variation yield diverse patterns of segmentation
among arthropods, patterns that have long been
studied by naturalists (see Bateson, 1894; Lankester,
1904). Indeed, standard patterns of segmentation
were so well known that in 1894 Bateson could
catalogue instances of exceptions found in nature.
These exceptions included a kind of variation in
which one segment in a series assumed the character
of another segment; Bateson called this phenomenon
‘‘homeosis.’’ The recognition many years later that
292
Annals of the
Missouri Botanical Garden
homeosis could be caused by mutations of a single
gene (Bridges & Morgan, 1923; Lewis, 1978) became
a springboard for thinking about how patterns of
segmentation might have evolved (Goldschmidt,
1940; King & Wilson, 1975; Lewis, 1978).
sequence of the transcription factors. (Bender et al.,
1983). These cis- regulatory regions scattered
throughout the remainder of the locus contain binding
sites for proteins that regulate the precise spatial and
temporal expression of Hox genes. Hox genes
function, in many animals, to pattern region-specific
cell fates along the body axis. They realize this
function by establishing specific expression domains
along the body axis and regulating large suites of
downstream target genes within these domains.
Precisely how this results in the ultimate body plan
of the animal is not completely known for any animal,
although it is best understood in Drosophila melanogaster. Given the understanding of the Hox gene
function in D. melanogaster, it was hypothesized that
arthropod segmental diversity would correlate with
changes in the regulation, both upstream and
downstream of the Hox genes (Grenier et al., 1997).
Evidence for intraspecific regulatory changes in Hox
gene expression domains and in Hox gene targets
followed in short order (for summary of regulatory
changes in Ubx, see Barton et al., 2007). Below, we
briefly recount some of the current data that support a
model of how shifting Hox boundaries and Hox targets
might explain how tagma evolved within crustaceans.
HYPOTHESES FROM DEVELOPMENTAL GENETICS
The discovery of the genetic basis of homeosis in
flies by Lewis (1978) laid the foundation for
understanding not only the developmental genetics
behind segmental patterning in the dipteran Drosophila melanogaster Meigen, but also the evolutionary diversification of segment character. Using
genetic analysis, the genes in the Antennapedia and
Bithorax complexes of D. melanogaster were shown to
control the development of the fruit fly body, with the
exception of the termini (Lewis, 1978; Kaufman et
al., 1980). The Hox genes have the unusual feature
that their order along the chromosome mirrors their
domains of function along the A-P body axis. This
chromosomal linearity also suggests deep ancestral
origins from multiple gene duplication events (Lewis,
1978).
Lewis discovered that the three posterior Hox
genes in Drosophila melanogaster define the limbless
abdominal body region. Given that the limbless
abdomen is a defining feature of hexapods, Lewis
(1978) speculated that the abdominal Hox genes
originated at the base of the insect lineage through a
serial gene duplication process. Arthropods with legs
on all their trunk segments were predicted to lack
these genes in their genomes, thereby avoiding the
repression of limb development in the posterior
region of their bodies. However, nearly all arthropods,
as well as their closest relatives, the Onychophora
Grube, or velvet worm phyllum, have a full
complement of Hox genes, and the simple hypothesis
correlating new Hox genes with arthropod diversification was abandoned (Grenier et al., 1997; see
below for discussion of why loss and gain of Hox
genes as a plausible genetic change underlying
arthropod diversification may yet be revived on a
smaller scale). The next hypotheses correlating the
evolution of tagma with Hox control of segment
identity were built with the knowledge that nearly all
arthropods have a full complement of Hox genes.
Once cloned, the Hox genes were identified as a
family of transcription factors (McGinnis et al., 1984;
Scott & Weiner, 1984) now recognized as a shared
feature of multicellular animals. Hox genes are found
throughout the Metazoa and are frequently clustered
in the genome (reviewed in Lemons & McGinnis,
2006). Interestingly, less than 5% of the Bithorax
locus identified by Lewis is devoted to the coding
EXISTING COMPARATIVE DATA
Crustaceans use the last three of their five head
appendages for feeding. However, in a number of
taxa, appendages on anterior thoracic segments have
been recruited to also function in feeding. For
example, decapods have eight thoracic segments
but only five pairs of locomotory limbs; the three
anterior thoracic segments are modified for feeding.
The first indication that the boundary between
feeding and nonfeeding thoracic limbs might be
under Hox control came from examining the
expression of Ubx protein in crustaceans with various
numbers of thoracic feeding limbs (Averof & Patel,
1997). Subsequent expression studies in isopod and
branchiopod crustaceans supported this hypothesis
(Abzhanov & Kaufman, 1999, 2000; Shiga et al.,
2002). More recently, it was demonstrated that RNA
interference (RNAi) silencing of Ubx in the peracarid
crustacean, Parhyale hawaiensis Dana, produces a
decrease in the gene expression of Ubx relative to
wildtype in the second and third thoracic segment.
This decrease in expression causes a transformation
of those limbs toward the feeding morphology of the
first thoracic appendage (Liubicich et al., 2009).
Conversely, ectopic expression of Ubx produces a
transformation of feeding appendages toward more
posterior limb morphologies (Pavopoulos et al.,
2009). These functional results are consistent with
Volume 99, Number 3
2014
Nagy & Williams
Morphological Transitions in Evolution
a model in which graded levels of Ubx protein control
the character of limbs along the thorax: high levels of
Ubx protein in the posterior thoracic limbs specify
thoracic identity, whereas lower levels in anterior
limbs specify a less elaborated, thoracic morphology.
These are compelling results and provide a plausible
model for re-specification and specialization of
anterior thoracic appendages during crustacean
evolution.
The paradigm of shifting boundaries is, however,
only useful for a subset of the morphological
transitions observed within the arthropods. Another
body of evidence is accruing that suggests changes
downstream of the Hox genes will play critical roles
in other morphological transitions. For example, in
flies, two Hox genes, Ubx and abdA, suppress limb
development in the abdomen by direct repression of
limb development genes (Vachon et al., 1992). The
crustacean Artemia Leach (brine shrimp) expresses
Ubx/abdA protein throughout the limb-bearing segments, but in Artemia, these genes do not appear to
repress limb development. Interestingly, the Artemia
Ubx gene is a weak repressor of the limb pathway,
and differences in the translational product or amino
acid sequence between Artemia and Drosophila
correlate well with their respective strength of
repression (Galant & Carroll, 2002; Ronshaugen et
al., 2002; Shiga et al., 2002). In addition, it appears
that the Artemia abdA mRNA is not translated into
protein (Hsia et al., 2010). Thus, while progressively
posterior boundaries of the Bithorax complex genes
are maintained in this crustacean, they do not share
the same regulatory targets as insects and do not
regulate boundaries in limb morphology. In sum, Hox
genes show a remarkable degree of conservation
throughout the evolution of the Metazoa; at the same
time, evolutionary changes in where they are
expressed and in their specific functions can help
explain evolutionary transitions in arthropod tagma.
arthropod body plan. For example, while the posterior
segments of the branchiopod crustacean, Artemia, do
not bear limbs, no Hox gene expression has been
detected in this region (Averof & Akam, 1995).
Secondly, although the vast majority of morphological
change occurs at boundaries between body tagma, in
some cases taxa differ mid-tagma, e.g., the collembolan furca, which appears in the middle of the
abdomen, from the fourth abdominal segment. (The
furca is a fused, forked appendage that gives the
name of springtails to collembolan insects.) This
region is not at an expected boundary of the Hox
genes and has not (as yet) been shown to be
associated with any novel boundaries of Hox genes.
This is related to another, much more common
phenomena not encompassed by this model. Numerous crustaceans have larval stages with patterns of
appendages quite distinct from their adult stages.
Specifically, they show differences in segmentation
patterns at sequential stages of the lifecycle that do
not consist of graded changes at boundaries.
Candidates for the genetic control of these morphological transitions are not yet obvious.
WHAT IT DOES AND DOES NOT EXPLAIN
Of the patterns of segmental variability to be
explained in arthropods, the Hox-based model
addresses one pattern very well: how segments within
a tagma are modified in a graded manner. The model
does not address changes in segment number in any
particular body region, i.e., how a taxon might evolve
from having a thorax with 12 to eight segments. This
is in part because the model focuses on limb
morphology, yet tagma are not defined just by limb
morphology but also by other segmental structures.
The comparative analysis of Hox genes has also
yielded the unexpected finding that in some species
Hox genes do not pattern nonterminal regions of the
293
SEGMENT NUMBER
VARIATION TO BE EXPLAINED
Segment number varies among the greater than
million species of arthropods, ranging from hundreds
in some millipedes (Enghoff et al., 1993) to eight in
ostracod crustaceans (Schram, 1986). Interestingly,
most classes and orders of arthropods do not vary in
segment number (with a few noteworthy exceptions,
e.g., the geophilomorph centipedes; Minelli &
Bortoletto, 1988), and segment number is a defining
character for some major lineages. Unfortunately, the
paradigm of using Drosophila melanogaster as a
starting point for hypotheses about the genetic/
developmental control of a morphological character
is not possible with the case of segment number.
Insects do not vary, for the most part, in total segment
number, and dipteran insects form their segments in
a highly derived manner. Mutations that increase
segment number have not been identified in D.
melanogaster. However, vertebrates show lineagespecific diversity in segment number and share with
arthropods the ancestral mode of adding segments
sequentially during development. Therefore, we use
models of segment development from vertebrates to
consider variation in segment number among arthropods. This comparison is supported not only by the
shared fact of sequential segment addition but also by
the finding that a number of the regulatory genes that
294
Annals of the
Missouri Botanical Garden
Figure 2. A–D. Variety in arthropod limb morphology, using crustacean examples. —A. Thoracic limb of the mysid shrimp,
Americamysis bahia Molenock. —B. Thoracic limb of the fairy shrimp, Thamnocephalus platyurus Packard. —C. Anterior
thoracic limb (maxilliped) of the grass shrimp, Paleomonetes pugio Holthuis. —D. Thoracic limb of the isopod, Cirolina
concharum Lat. The two main branches are labeled 1 and 2, with additional lobes labeled as medial (M) or lateral (L). Note that
beyond the fundamental variability in number of branches and lobes, limb parts are highly variable in terms of shape,
proportion, and setal numbers and morphology. —E. Patterning in the leg disc of Drosophila. Diagram at top indicates signaling
along the A-P segment boundary, which initiates PD outgrowth of the leg. The genes that establish the PD axis—Distal-less
(red), dachshund (green), and extradenticle/homothorax (blue)—are activated in circular domains in the larval leg disc by a
combination of signals (gold, light blue). As the larval leg disc grows and extends into the adult leg, these genes function to
pattern three domains along the PD axis of the leg (diagram at bottom). Experimentally initiating new PD outgrowths by
misexpression of a signal (gold) gives rise to artificially branched legs. F, G. Comparative expression data in limbs of varying
Volume 99, Number 3
2014
Nagy & Williams
Morphological Transitions in Evolution
operate in vertebrate segmentation play similar roles
in arthropods (Peel et al., 2005).
more numerous somites. Do arthropods similarly
control the size of segments generated per cell
generation in the region of the embryo that forms
segments to control segment number? As we point out
below, understanding of the fundamental cellular
processes of segmentation in arthropods lags behind
that of vertebrates. Consequently, understanding the
genetic control of segment number in arthropods is
only just emerging.
HYPOTHESES FROM DEVELOPMENTAL GENETICS
In vertebrates, segments arise from embryonic
somites, which develop sequentially in a head-to-tail
direction in the embryo (Fig. 1D). Somites bud off
from the anterior presomitic mesoderm, an unpatterned region of active growth in the posterior of the
embryo. Somites form at a species-specific rate, e.g.,
30 min./segment in zebrafish, 90 min./segment in
chickens, and 120 min./segment in mice (Romanoff,
1960; Tam, 1981; Schröter et al., 2008). At the same
time, cells are added to the posterior of the presomitic
mesoderm through the process of gastrulation,
thereby allowing for continued development of
segments. The size and number of somites depend
on a dynamic interaction between three factors: the
size of the presomitic mesoderm, the position of a
posteriorly moving wavefront of determination, and
oscillations of certain genes known as the segmentation clock (Fig. 1E). In general, in those vertebrate
species examined, all use a similar molecular toolkit
to run the segmentation clock: mainly genes of the
Notch, FGF, and Wnt signaling pathways. In mice
and fish, mutations in the oscillator genes cause
severe defects in the somites. These pathways
function to make pulses of signaling molecules in
the presomitic mesoderm (Cooke & Zeeman, 1976;
Elsdale et al., 1976; Palmeirim et al., 1997; Dubrulle
et al., 2001; Sawada et al., 2001). Each cycle of the
oscillator converts oscillations in time to a periodic
pattern in space and results in the appearance of a
pair of segments.
Gomez et al. (2008) asked whether the evolutionary variation in segment number between snakes,
mice, chicken, and zebrafish could have resulted
from developmentally varying either the size of the
presomitic mesoderm, the position of the determination wavefront, or the periodicity of the segmentation
clock. They found that in snakes, which have a high
number of segments, the rate of oscillation of the
segmentation clock was high relative to the growth
and elongation of the presomitic mesoderm. Thus,
snake embryos segment the presomitic mesoderm
faster than other vertebrates, making smaller and
295
EXISTING COMPARATIVE DATA
In most arthropods, segments form in an A-P
progression (Sander, 1976; Minelli & Fusco, 2004;
Peel et al., 2005). However, there are surprisingly few
data that indicate whether segments form with a
species-specific periodicity since patterns of segment
addition are typically described only with reference to
morphological stage and not developmental time. The
assumption is that segment addition is regular and, in
examining some crustaceans, we have found that
segments are added with linear periodicity (Williams
et al., 2012). Whether a regular periodicity in
segment addition is widespread in arthropods
remains unknown.
The combination of three features that control
sequential segment addition in vertebrates—a growth
zone, a determination wavefront, and a segmentation
clock—has not been demonstrated for any arthropod.
In general, most sequentially segmenting arthropods
have a region of unpatterned tissue in the posterior
that generates segments, i.e., a growth zone. However,
the extent of the unpatterned tissue and its rate of
growth or depletion during the process of segment
addition are completely unknown. There is no
evidence as of yet for a determination front in
arthropods, at least of the kind found in vertebrates
that is regulated by antagonistic gradients of signaling
molecules. However, a growing body of literature
suggests that the molecular toolkit that runs the
vertebrate segmentation clock is conserved in
spiders, sequentially segmenting insects, and recently, we have found evidence for the function of clock
orthologs in crustaceans (Williams et al., 2012). In
each of these cases, Notch signaling has been
demonstrated to play a role in the proper formation
of sequentially added segments.
morphologies. —F. Simplified schematic of gene expression in limbs of two crustaceans. The genes that establish the PD axis
include Distal-less (red), dachshund (yellow), and extradenticle/homothorax (green) and are expressed in a pattern similar to
Drosophila (at left) and Porcellio scaber Latreille (at right). —G. In Triops longicaudatus LeConte, the three genes are expressed
even in the unusually shaped limb bud of this species that develops into a highly modified (phyllopodous) limb form. The
asterisk marks the two distal branches in the schematic of the limb bud and the adult limb.
296
Annals of the
Missouri Botanical Garden
WHAT IT DOES AND DOES NOT EXPLAIN
HYPOTHESES FROM DEVELOPMENTAL GENETICS
The discovery that Notch signaling plays a role in
sequentially segmenting arthropods was initially
hailed to indicate that vertebrate and arthropod
segmentation was homologous. Closer comparison
shows that, even among arthropods, disruption of the
Notch signaling network has variable effects. In some
species like Drosophila melanogaster, mutations in
Notch signaling have no consequences for segmentation. While the discovery of a role for Notch
signaling in arthropod segmentation is significant, a
robust model of Notch function as well as the
possibility that Notch signaling serves as a molecular
oscillator remains unresolved. Thus, whether the
segment number is regulated in arthropods via a
balance between clock rate and rate of growth in the
posterior awaits further research.
In Drosophila melanogaster, limb primordia are
positioned at the boundary that defines the posterior
portion or compartment of each segment. Subsequently, via signaling activated at the A-P compartment boundary, limb axes are defined and proximodistal (PD) elongation occurs. The gene network
involved in patterning the PD leg axis is well
described (Fig. 2E–G; reviewed in Nagy & Williams,
2001; Angelini & Kaufman, 2005b). In short, the leg
is divided into three domains along the PD axis and
patterned by genes with mutually exclusive gene
expression domains. Loss of these gene expression
domains causes loss of position-specific leg tissue
and truncated or shortened legs. Experiments in D.
melanogaster demonstrated that it was possible to
partially duplicate the PD axis; manipulating signaling along the A-P boundary formed new sites of PD
elongation and ultimately branched legs with duplicated distal axes (Struhl & Basler, 1993; DiazBenjumea et al., 1994). This led to the hypothesis
that reiterating the PD patterning network along the
A-P segment boundary could have generated naturally occurring, branched limbs (Campbell & Tomlinson, 1995).
LIMB MORPHOLOGY
VARIATION TO BE EXPLAINED
The array of limb structures in arthropods is truly
astounding (Fig. 2A–D; Brusca & Brusca, 2003).
Even apparently simple, cylindrical limbs, like the
walking legs of a crab, may have elaborate lateral
outgrowths (in this case, functioning as gills and
hidden beneath the carapace). Beyond cylindrical
walking legs, arthropod limbs show adaptations for
swimming, grasping, sensing, food handling, and
many other functions. Correspondingly, limbs may be
flattened into paddles, calcified for crushing pincers,
or adorned with elaborate setal arrays. It is tempting
to organize all arthropod limbs as variations on a
theme of a single limb axis with medial or lateral
outgrowths. However, it is not clear that this
characterization is evolutionarily accurate, since
some would argue that the ancestral limb had two
fundamental branches (Walossek, 1993, 1999;
Boxshall, 2004). In one major arthropod group, in
extant Crustacea, the limb has two branches, i.e., the
main axis is bifurcated. Thus, to understand some of
the variation in limb morphology, there are at least
three main questions to be addressed: (1) what
patterns the main axial outgrowth; (2) how is the main
limb axis bifurcated; and (3) what patterns the vast
array of medial and lateral outgrowths that occur
proximally on the limb axis? In addition to these main
categories of variation, there are numerous features of
appendages that differ widely between limbs, both
within and between taxa, e.g., setal type and number,
cuticular thickness and specialization or jointing, etc.
This variation is often crucial to functional differences between limbs and also needs to be explained.
EXISTING COMPARATIVE DATA
Based on the hypothesis above, limb patterning
genes from Drosophila melanogaster were candidates
to regulate limb patterning in other species. When D.
melanogaster genes were examined in other species,
both in expression and function, it became clear that,
while some genes are expressed similarly across
arthropods, the entire network from D. melanogaster
was not conserved. Critically, the wingless and
decapentaplegic genes that function directly upstream
of PD elongation in D. melanogaster do not show
conserved function, even within insects, and so
cannot explain modulation of PD elongation (Angelini
& Kaufman, 2005a, 2005b). PD elongation is based
primarily on the activation of Distal-less in D.
melanogaster. Although analyses of Distal-less function in other arthropods show that it is required for
PD growth of limbs (Beerman et al., 2001; Shoppmeier & Damen, 2001; Khila & Grbic, 2007), there is
no evidence that the PD patterning network is
reiterated to form branches. Instead, in every case
where it has been examined, the evidence points to a
single PD patterning axis whether the limb has only
one axis, a bifurcated axis, or is a highly modified
paddle (Fig. 2G; Williams, 1998; reviewed in
Williams & Nagy, 2001; Williams et al., 2002).
Volume 99, Number 3
2014
Nagy & Williams
Morphological Transitions in Evolution
Indeed, the network of genes regulating PD outgrowth
is broadly conserved.
far has the paradigm of diversification using a finite
toolkit taken us within the field of arthropod
EvoDevo? With respect to the three areas analyzed
herein, each feature shows surprising instances of
deep conservation of certain patterning mechanisms.
For tagmatization and limb identity along the A-P
axis, Hox genes play a fundamental role in shifting
the boundaries of limb morphology. For variation in
segment number, Notch involvement is widespread.
For limb morphology, there is a highly conserved PD
patterning module.
The idea of a finite toolkit has proven to be a
surprisingly robust hypothesis. In many cases, when
we examine nonmodel organisms, we find the same
regulatory genes in the same roles they play in model
systems. Furthermore, some aspects of variability can
be explained by changes in broadly conserved genes.
For example, the subspecialization of limbs within
tagma seems to be well modeled by shifts in the
boundaries of Hox expression. However, what has
also become clear is that there are two main
drawbacks to following this approach. First, candidate genes may not be widely conserved or may have
pleiotropic effects that complicate our modeling their
roles. Second, and more profound, the level of
patterning that is deeply conserved is often not the
level that establishes the details of morphology that
are fundamental to adaptive radiations. This is most
clearly illustrated by the analysis of limbs. Whereas
there does seem to be a widespread PD patterning
module that forms a single PD axis in all limbs, we
have no general models to explain the morphological
variability that characterizes the functional diversification of limbs. Nor do we know how the limb axis is
bifurcated in two-branched forms. Given that the
branching and outgrowths of limbs are the very
substrate of their functional diversity, this conserved
patterning module has not served us well in analyzing
that diversification.
Even as we write this, EvoDevo is being
transformed by the development of new tools that in
part address some of the limitations. Transgenesis
and RNAi are transforming our ability to conduct
functional studies in nonmodel arthropods. These
methods will facilitate a much-needed wider taxon
sampling. With RNAi, we can test gene function in
nonmodel systems and test whether the assumptions
derived from a few model systems hold more broadly.
High-throughput sequencing/proteomics speed up
gene discovery at an incredible rate and are limited
only by financial resources rather than the life history
strategies of an organism. The sequenced genomes
confirm that the generic developmental toolkit is
widely present. Interestingly, 30%–60% of open
WHAT IT DOES AND DOES NOT EXPLAIN
Two points in limb development appear broadly
conserved with little variation. First, all limbs
examined are positioned along the A-P segment
boundary. In spite of this conserved positioning, the
signaling that subsequently occurs along the A-P
boundary that initiates PD outgrowth in Drosophila
melanogaster is not conserved. Nevertheless, once it
is initiated, the network of PD leg patterning and
elongation is the second broadly conserved aspect of
leg patterning. This appears to be the case even in
limbs of highly divergent morphology, like the
flattened, multilobed paddles of branchiopod crustaceans. The deep conservation of PD patterning is
striking and probably represents a core set of genes
that, once activated, can produce a limb (i.e., a limbpatterning module). However, one core aspect of
variation in limbs, branches or outgrowths from the
main axis, is not explained by the comparative data.
The analysis of candidate genes yielded no patterns
that gave rise to new hypotheses explaining branching, and we currently have no good working models to
account for such limb variation.
The analysis of candidate genes from Drosophila
melanogaster limb patterning is complicated by the
fact that a number of genes involved in limb
patterning have pleiotropic effects. For example,
Distal-less protein is found in almost every appendage
examined to date, but evaluating its role in patterning
limb outgrowth is confounded by its additional role in
sensory development (Mittmann & Scholtz, 2001;
Williams et al., 2002; Williams, 2008). Although
Distal-less is well known to function in the nervous
system of D. melanogaster (Panganiban, 2000), its
role in limb patterning is distinct both spatially and
temporally because of D. melanogaster’s specialized
metamorphic mode of development, where the
segregation and patterning of cells fated to become
limbs occur much earlier than the differentiation of
limb sensory structures. Most arthropods lack this
segregation between limb patterning and limb
differentiation, and, therefore, gene expression regulating limb patterning overlaps gene expression
regulating sensory patterning, confounding the inference of function in limbs with complex morphology.
CONCLUSIONS
EvoDevo promised to revive and answer some
longstanding questions about the morphological
diversity that results from adaptive radiations. How
297
298
Annals of the
Missouri Botanical Garden
reading frames identified in the arthropod species
sequenced to date have no identifiable orthologs in
other species, and as much as 50% of the active
transcriptome in some species are these orphan
genes. These orphan genes may have important
functions in nonmodel systems. These new approaches are opening new windows on how genomes evolve.
How directly those windows will lead to more a
complex understanding of the kinds of changes that
natural selection can act upon remains to be seen.
In addition to the increased use of genomics,
transcriptomics, and proteomics to further our
mechanistic understanding of development, several
other parallel trends or research programs can be
discerned (e.g., Müller, 2007). One is a trend toward
using computational and bioinformatics approaches
to understand development and how developmental
processes evolve; another is the integration of
ecological and environmental aspects of developmental biology into what is often called EcoDevo (Gilbert,
2001). In addition, evolution and development
studies applied to smaller-scale evolutionary problems within insects have had success (e.g., Kopp,
2011; Prud’homme et al., 2011; Frankel et al., 2012).
Finally, it is sometimes forgotten that another
important trend is continued emphasis, especially
in marine invertebrates (Love, 2009), on developing a
phylogenetically informed comparative embryology.
We conclude that EvoDevo has made great strides
in uncovering the common features underlying the
development of morphology. The field has also made
a start at the discovery of both protein and regulatory
changes that correlate with morphological changes.
At the same time, resolution of many longstanding
questions about morphological diversity has largely
not occurred. Arthropod morphological diversity is
structurally complex and not often captured by
broadly conserved patterning elements. The ultimate
goal is to understand the complete arc from genetic
regulation to expressed phenotype in a set of related
organisms. This goal of understanding how the
genotype translates into the complex, evolvable
morphological phenotype remains for the future.
(Hemiptera) support a role for Wnt signaling in body
segmentation but not appendage development. Developm. Biol. 283: 409–423.
Angelini, D. R. & T. C. Kaufman. 2005b. Insect
appendages and comparative ontogenetics. Developm.
Biol. 286 57–77.
Appel, T. 1987. The Cuvier-Geoffroy Debate: French
Biology in the Decades Before Darwin. Oxford University
Press, Oxford.
Averof, M. & M. Akam. 1995. Hox genes and the
diversification of insect and crustacean body plans.
Nature 376: 420–423.
Averof, M. & N. H. Patel. 1997. Crustacean appendage
evolution associated with changes in Hox gene expression. Nature 388: 682–686.
Barton, N. H., D. E. G. Briggs, J. A. Eisen, D. B. Goldstein
& N. H. Patel. 2007. Evolution. Cold Spring Harbor
Press, New York.
Bateson, W. 1894. Materials for the Study of Variation
Treated with Especial Regard to Discontinuity in the
Origin of Species. Macmillan, London.
Beermann, A., D. G. Jay, R. W. Beeman, M. Hülskamp, D.
Tautz & G. Jürgens. 2001. The short antennae gene of
Tribolium is required for limb development and encodes
the orthologue of the Drosophila Distal-less protein.
Development 128: 287–297.
Bender, W., M. Akam, F. Karch, P. A. Beachy, M. Peifer, P.
Spierer, E. B. Lewis & D. S. Hogness. 1983. Molecular
genetics of the Bithorax complex in Drosophila melanogaster. Science 221: 23–29.
Bowler, P. J. 2003. Evolution: The History of an Idea, 3rd
ed. University of California Press, Berkeley.
Boxshall, G. A. 2004. The evolution of arthropod limbs.
Biol. Rev. Cambridge Philos. Soc. 79: 253–300.
Bridges, C. B. & T. H. Morgan. 1923. The ThirdChromosome Group of Mutant Characters of Drosophila
melanogaster. Carnegie Institution of Washington, Washington, D.C.
Brusca, R. C. & G. J. Brusca. 2003. Invertebrates, 2nd ed.
Sinauer Associates, Sunderland, Massachusetts.
Campbell, G. & A. Tomlinson. 1995. Initiation of
proximodistal axis in insect legs. Development 121:
619–628.
Carroll, S. B. 2005. Endless Forms Most Beautiful: The New
Science of EvoDevo and the Making of the Animal
Kingdom. W. W. Norton & Co., New York.
Carroll, S. B., J. K. Grenier & S. D. Weatherbee. 2001.
From DNA to Diversity: Molecular Genetics and the
Evolution of Animal Design. Blackwell Publishing,
Malden, Massachusetts.
Cooke, J. & E. C. Zeeman. 1976. A clock and wavefront
model for control of the number of repeated structures
during animal morphogenesis. J. Theor. Biol. 58: 455–
476.
De Robertis, E. M. & Y. Sasai. 1996. A common plan for
dorsoventral patterning in Bilateria. Nature 380: 37–40.
Dequéant, M. L. & O. Pourquié. 2008. Segmental
patterning of the vertebrate embryonic axis. Nat. Rev.
Genet. 9: 370–382.
Diaz-Benjumea, F. J., B. Cohen & S. M. Cohen. 1994. Cell
interactions between compartments establishes the
proximal-distal axis of Drosophila legs. Nature 372:
175–179.
Dubrulle, J., M. J. McGrew & O. Pourquié. 2001. FGF
signaling controls somite boundary position and regulates
Literature Cited
Abzhanov, A. & T. C. Kaufman. 1999. Novel regulation of
the homeotic gene Scr associated with a crustacean legto-maxilliped appendage transformation. Development
126: 1121–1128.
Abzhanov, A. & T. C. Kaufman. 2000. Embryonic
expression patterns of the Hox genes of the crayfish
Procambarus clarkii (Crustacea, Decapoda). Evol. Developm. 2: 271–283.
Angelini, D. R. & T. C. Kaufman. 2005a. Functional
analyses in the milkweed bug Oncopeltus fasciatus
Volume 99, Number 3
2014
Nagy & Williams
Morphological Transitions in Evolution
299
segmentation clock control of spatiotemporal Hox gene
activation. Cell 106: 219–232.
Elsdale, T., M. Pearson & M. Whitehead. 1976. Abnormalities in somite segmentation following heat shock to
Xenopus embryos. J. Embryol. Exp. Morphol. 35: 625–
635.
Enghoff, H., W. Dohle & J. G. Blower. 1993. Anamorphosis
in millipedes (Diplopoda)—The present state of knowledge with some developmental and phylogenetic considerations. Zool. J. Linn. Soc. 109: 103–234.
Frankel, N., W. Wang & D. L. Stern. 2012. Conserved
regulatory architecture underlies parallel genetic changes
and convergent phenotypic evolution. Proc. Natl. Acad.
Sci. U.S.A. 109: 20,975–20,979.
Galant, R. & S. B. Carroll. 2002. Evolution of a
transcriptional repression domain in an insect Hox
protein. Nature 415: 910–913.
Gehring, W. J. 2002. The genetic control of eye
development and its implications for the evolution of
the various eye-types. Int. J. Developm. Biol. 46: 65–73.
Geoffroy Saint-Hilaire, E. 1822. Considérations générales
sur la vertèbre. Mém, Mus. Hist. Nat. 9: 89–119, plate 7
[p. 119].
Gilbert, S. F. 2001. Ecological developmental biology:
Developmental biology meets the real world. Developm.
Biol. 233: 1–12.
Gilbert, S. F., J. M. Opitz & R. A. Raff. 1996.
Resynthesizing evolutionary and developmental biology.
Developm. Biol. 173: 357–372.
Goldschmidt, R. 1940. The Material Basis of Evolution.
Yale University Press, New Haven.
¨
Gomez, C., E. M. Ozbudak,
J. Wunderlich, D. Baumann, J.
Lewis & O. Pourquié. 2008. Control of segment number
in vertebrate embryos. Nature 454: 335–339.
Grenier, J., T. Garber, R. Warren, P. Whitington & S.
Carroll. 1997. Evolution of the entire arthropod Hox gene
set predated the origin and radiation of the onychophoran/arthropod clade. Curr. Biol. 7: 547–553.
Hall, B. K. 1994. Homology: The Hierarchical Basis of
Comparative Biology. Academic Press, San Diego.
Hall, B. K. 1999. Evolutionary Developmental Biology.
Kluwer Academic Publishers, Dordrecht.
Hsia, C., A. Paré, M. Hannon, M. Ronshaugen & W.
McGinnis. 2010. Silencing of an abdominal Hox gene
during early development is correlated with limb
development in a crustacean trunk. Evol. Developm.
12: 131–143.
Jacob, F. 1977. Evolution and tinkering. Science 196:
1161–1166.
Kaufman, T. C., R. Lewis & B. Wakimoto. 1980.
Cytogenetic analysis of chromosome 3 in Drosophila
melanogaster: The homoeotic gene complex in polytene
chromosome interval 84A–B. Genetics 94: 115–133.
Khila, A. & M. Grbic. 2007. Gene silencing in the spider
mite Tetranychus urticae: dsRNA and siRNA parental
silencing of the Distal-less gene. Developm. Genes Evol.
217: 241–251.
King, M. C. & A. C. Wilson. 1975. Evolution at two levels
in humans and chimpanzees. Science 188: 107–116.
Kopp, A. 2011. Drosophila sex combs as a model of
evolutionary innovations. Evol. Developm. 13: 504–522.
Lankester, E. R. 1904. The structure and classification of
Arthropoda. Quart. J. Microscop. Sci. London 47: 523–
582.
Lemons, D. & W. McGinnis. 2006. Genomic evolution of
Hox gene clusters. Science 313: 1918–1922.
Lewis, E. B. 1978. A gene complex controlling segmentation in Drosophila. Nature 276: 565–570.
Liubicich, D. M., J. M. Serano, A. Pavlopoulos, Z.
Kontarakis, M. E. Protas, E. Kwan, S. Chatterjee, K. D.
Tran, M. Averof & N. H. Patel. 2009. Knockdown of
Parhyale Ultrabithorax recapitulates evolutionary changes in crustacean appendage morphology. Proc. Natl.
Acad. Sci. U.S.A. 106: 13,892–13,896.
Love, A. C. 2009. Marine invertebrates, model organisms,
and the modern synthesis: Epistemic values, evo-devo,
and exclusion. Theory Biosci. 128: 19–42.
Mayr, E. 1993. One Long Argument: Charles Darwin and
the Genesis of Modern Evolutionary Thought. Harvard
University Press, Cambridge.
McGinnis, W., M. S. Levine, E. Hafen, A. Kuroiwa & W. J.
Gehring. 1984. A conserved DNA sequence in homoeotic
genes of the Drosophila Antennapedia and Bithorax
complexes. Nature 308: 428–433.
Minelli, A. & S. Bortoletto. 1988. Myriapod metamerism
and arthropod segmentation. Biol. J. Linn. Soc. 33: 323–
343.
Minelli, A. & G. Fusco. 2004. Evo-devo perspectives on
segmentation: Model organisms, and beyond. Trends
Ecol. Evol. 19: 423–429.
Mittmann, B. & G. Scholtz. 2001. Distal-less expression in
embryos of Limulus polyphemus (Chelicerata, Xiphosura)
and Lepisma saccharina (Insecta, Zygentoma) suggests a
role in the development of mechanoreceptors, chemoreceptors, and the CNS. Developm. Genes Evol. 211: 232–
243.
Müller, G. B. 2007. EvoDevo: Extending the evolutionary
synthesis. Nat. Rev. Genet. 8: 943–950.
Nagy, L. M. & T. A. Williams. 2001. Comparative limb
development as a tool for understanding the evolutionary
diversification of the limbs in arthropods: Challenging the
modularity paradigm. Pp. 455–488 in G. P. Wagner
(editor), The Character Concept in Evolutionary Biology.
Academic Press, San Diego.
Palmeirim, I., D. Henrique, D. Ish-Horowicz & O. Pourquié.
1997. Avian hairy gene expression identifies a molecular
clock linked to vertebrate segmentation and somitogenesis. Cell 91: 639–648.
Panganiban, G. 2000. Distal-less function during Drosophila
appendage and sense organ development. Developm.
Dynam. 218: 554–562.
Pavlopoulos, A., Z. Kontarakis, D. M. Liubicich, J. M.
Serano, M. Akam, N. H. Patel & M. Averof. 2009.
Probing the evolution of appendage specialization by Hox
gene misexpression in an emerging model crustacean.
Proc. Natl. Acad. Sci. U.S.A. 106: 13,897–13,902.
Peel, A. D., A. D. Chipman & M. Akam. 2005. Arthropod
segmentation: Beyond the Drosophila paradigm. Nat.
Rev. Genet. 6: 905–916.
Prud’homme, B., C. Minervino, M. Hocine, J. D. Cande, A.
Aouane, H. D. Dufour, V. A. Kassner & N. Gompel.
2011. Body plan innovation in treehoppers through the
evolution of an extra wing-like appendage. Nature 473:
83–86.
Romanoff, A. 1960. The Avian Embryo: Structural and
Functional Development. Macmillan Co., New York.
Ronshaugen, M., N. McGinnis & W. McGinnis. 2002. Hox
protein mutation and macroevolution of the insect body
plan. Nature 415: 914–917.
Russell, E. S. 1917. Form and Function: A Contribution to
the History of Animal Morphology. Dutton and Co., New
York.
300
Annals of the
Missouri Botanical Garden
Sander, K. 1976. Specification of the basic body pattern in
insect embryogenesis. Advances Insect Physiol. 12: 125–
238.
Sawada, A., M. Shinya, Y. J. Jiang, A. Kawakami, A.
Kuroiwa & H. Takeda. 2001. Fgf/MAPK signaling is a
crucial positional cue in somite boundary formation.
Development 128: 4873–4880.
Schoppmeier, M. & W. G. Damen. 2001. Double-stranded
RNA interference in the spider Cupiennius salei: The role
of Distal-less is evolutionarily conserved in arthropod
appendage formation. Developm. Genes Evol. 211: 76–
82.
Schram, F. R. 1986. Crustacea. Oxford University Press,
Oxford.
Schröter, C., L. Herrgen, A. Cardona, G. J. Brouhard, B.
Feldman & A. C. Oates. 2008. Dynamics of zebrafish
somitogenesis. Developm. Dynam. 237: 545–553.
Scott, M. & A. Weiner. 1984. Structural relationships
among genes that control development sequence homology between Antennapedia, Ultrabithorax and fushi
tarazu loci of Drosophila. Proc. Natl. Acad. Sci. U.S.A.
81: 4115–4119.
Shiga, Y., R. Yasumoto, H. Yamagata & S. Hayashi. 2002.
Evolving role of Antennapedia protein in arthropod limb
patterning. Development 129: 3555–3561.
Struhl, G. & K. Basler. 1993. Organizing activity of wingless
protein in Drosophila. Cell 72: 527–540.
Tam, P. P. 1981. The control of somitogenesis in mouse
embryos. J. Embryol. Exp. Morphol. 65: 103–128.
Vachon, G., B. Cohen, C. Pfeifle, M. McGuffin, J. Botas &
S. M. Cohen. 1992. Homeotic genes of the Bithorax
complex repress limb development in the abdomen of the
Drosophila embryo through the target gene Distal-less.
Cell 71: 437–450.
Vonk, F. J. & M. K. Richardson. 2008. Developmental
biology: Serpent clocks tick faster. Nature 454: 282–283.
Walossek, D. 1993. The Upper Cambrian Rehbachiella
kinnekullensis and the phylogeny of Branchiopoda and
Crustacea. Fossils Strata 32: 1–202.
Walossek, D. 1999. On the Cambrian diversity of
Crustacea. Pp. 3–27 in F. R. Schram & J. C. von Vaupel
Klein (editors), Crustaceans and the Biodiversity Crisis:
Proceedings of the Fourth International Crustacean
Congress. Brill, Leiden.
Williams, T. A. 1998. Distalless expression in crustaceans
and the patterning of branched limbs. Developm. Genes
Evol. 207: 427–434.
Williams, T. A. 2008. Early Distal-less expression in a
developing crustacean limb bud becomes restricted to
setal forming cells. Evol. Developm. 10: 114–120.
Williams, T. A. & L. M. Nagy. 2001. Developmental
modularity and the evolutionary diversification of
arthropod limbs. J. Exp. Zool. 291: 241–257.
Williams, T. A., C. Nulsen & L. M. Nagy. 2002. A complex
role for Distal-less in crustacean appendage development.
Developm. Biol. 241: 302–312.
Williams, T. A., B. Blachuta, T. A. Hegna & L. M. Nagy.
2012. Decoupling elongation and segmentation: Notch
involvement in anostracan crustacean segmentation.
Evol. Developm. 14: 372–382.